[Pacemaker] ocf:heartbeat:Filesystem doesn't work via corosync

[Dan Frincu]

Hi,

On Thu, Aug 18, 2011 at 12:59 AM, Cotton Tenney
<cotton.tenney at rogerspos.com> wrote:
>  Hello all,
>
> I've been trying to get a 2-node mysql cluster up on Pacemaker/DRBD and I'm having issues getting corosync to mount the partition after starting DRBD. It seems like when running the ocf:Filesystem script via pacemaker, it fails to make the drbd device primary on the master node. I can manually run the ocf Filesystem, and DRBD starts and the partition is mounted as expected:
>
> root at gila:/usr/lib/ocf/resource.d/heartbeat# export OCF_ROOT=/usr/lib/ocf
> root at gila:/usr/lib/ocf/resource.d/heartbeat# export OCF_RESKEY_device="/dev/drbd0"
> root at gila:/usr/lib/ocf/resource.d/heartbeat# export OCF_RESKEY_options=rw
> root at gila:/usr/lib/ocf/resource.d/heartbeat# export OCF_RESKEY_fstype=ext4
> root at gila:/usr/lib/ocf/resource.d/heartbeat# export OCF_RESKEY_directory="/var/lib/mysql"
> root at gila:/usr/lib/ocf/resource.d/heartbeat# /etc/init.d/drbd start
>  --== Thank you for participating in the global usage survey ==--
> The server's response is:
> node already registered
> * Starting DRBD resources
> [
> disk0
> Found valid meta data in the expected location, 320000978944 bytes into /dev/fioa.
> d(disk0) s(disk0) n(disk0) ] ...done.
> root at gila:/usr/lib/ocf/resource.d/heartbeat# drbdadm primary all
> root at gila:/usr/lib/ocf/resource.d/heartbeat# ./Filesystem start
> Filesystem[28277]: INFO: Running start for /dev/drbd0 on /var/lib/mysql
> FATAL: Module scsi_hostadapter not found.
> Filesystem[28277]: INFO: Starting filesystem check on /dev/drbd0
> fsck from util-linux-ng 2.17.2
> /dev/drbd0: clean, 79439/19537920 files, 7781765/78122846 blocks
> root at gila:/usr/lib/ocf/resource.d/heartbeat#
>
>
> My config:
>
> root at gila:~# crm configure show
> node gila
> node kanab
> primitive drbd_mysql ocf:linbit:drbd \
> params drbd_resource="disk0" \
> op monitor interval="15s" \
> op start interval="15s"
> primitive fs_mysql ocf:heartbeat:Filesystem \
> params device="/dev/drbd0" directory="/var/lib/mysql" fstype="ext4"
> op monitor interval="15s" \
> op start interval="15s"
> primitive ip_mysql ocf:heartbeat:IPaddr2 \
> params ip="192.168.10.101" nic="eth0"
> primitive mysqld lsb:mysql
> group mysqlgroup fs_mysql ip_mysql mysqld
> ms ms_drbd_mysql drbd_mysql \
> meta master-max="1" master-node-max="1" clone-max="2" clone-node-max="1" notify="true"
> colocation mysql_on_drbd inf: mysqlgroup ms_drbd_mysql:Master

You need an ordering constraint specifying start mysqlgroup after DRBD
gets promoted to Master.
order mysql_after_drbd inf: ms_drbd_mysql:promote mysqlgroup:start

Regards,
Dan

> property $id="cib-bootstrap-options" \
> dc-version="1.0.8-042548a451fce8400660f6031f4da6f0223dd5dd" \
> cluster-infrastructure="openais" \
> expected-quorum-votes="1" \
> stonith-enabled="false"
> root at gila:~# cat /etc/drbd.d/disk0.res
> resource disk0 {
> protocol C;
> net {
> cram-hmac-alg sha1;
> shared-secret "xxx";
> }
> on kanab {
> device /dev/drbd0;
> disk /dev/fioa;
> address 10.10.1.2:7788;
> meta-disk internal;
> }
> on gila {
> device /dev/drbd0;
> disk /dev/fioa;
> address 10.10.1.1:7788;
> meta-disk internal;
> }
> }
> root at gila:~#
>
> Log messages:
>
> Aug 17 12:35:20 gila crmd: [24757]: info: te_rsc_command: Initiating action 32: start fs_mysql_start_0 on gila (local)
> Aug 17 12:35:20 gila crmd: [24757]: info: do_lrm_rsc_op: Performing key=32:1:0:9b90831c-824b-462f-a2db-7e582c706638 op=fs_mysql_start_0 )
> Aug 17 12:35:20 gila lrmd: [24754]: debug: on_msg_perform_op:2359: copying parameters for rsc fs_mysql
> Aug 17 12:35:20 gila lrmd: [24754]: debug: on_msg_perform_op: add an operation operation start[8] on ocf::Filesystem::fs_mysql for client 24757, its parameters: fstype=[ext4] crm_feature_set=[3.0.1] device=[/dev/drbd0] CRM_meta_timeout=[20000] options=[rw] directory=[/var/lib/mysql] to the operation list.
> Aug 17 12:35:20 gila lrmd: [24754]: info: rsc:fs_mysql:8: start
> Aug 17 12:35:20 gila Filesystem[25276]: INFO: Running start for /dev/drbd0 on /var/lib/mysql
> Aug 17 12:35:20 gila lrmd: [24754]: info: RA output: (fs_mysql:start:stderr) FATAL: Module scsi_hostadapter not found.
> Aug 17 12:35:20 gila Filesystem[25276]: INFO: Starting filesystem check on /dev/drbd0
> Aug 17 12:35:20 gila lrmd: [24754]: info: RA output: (fs_mysql:start:stdout) fsck from util-linux-ng 2.17.2
> Aug 17 12:35:20 gila lrmd: [24754]: info: RA output: (fs_mysql:start:stdout) Disk write-protected; use the -n option to do a read-only#012check of the device.
> Aug 17 12:35:20 gila lrmd: [24754]: info: RA output: (fs_mysql:start:stderr) fsck.ext4: Read-only file system while trying to open /dev/drbd0#015
>
>
> Any help would be greatly appreciated.
>
> Thanks,
> Cotton Tenney
> Systems Administrator
> Rogers Software Development
>
>
> _______________________________________________
> Pacemaker mailing list: Pacemaker at oss.clusterlabs.org
> http://oss.clusterlabs.org/mailman/listinfo/pacemaker
>
> Project Home: http://www.clusterlabs.org
> Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
> Bugs: http://developerbugs.linux-foundation.org/enter_bug.cgi?product=Pacemaker
>



-- 
Dan Frincu
CCNA, RHCE